sysusers prep patches #1763
sysusers prep patches #1763
Conversation
This is easy to set in the environment for local development and avoid leaking space in the success cases.
Will be used by the sysusers code to pass down a pipe that the child writes to.
We should only be calling into GIO once we've set our preparatory environment variables etc. Not fixing anything today that I know of, just noticed while looking at the code.
There was a problem hiding this comment.
None of the RPM packages I can find do this.
Hmm, just looking at the "a"s, I see:
abrt.spec:909:%dir %attr(0751, root, abrt) %{_localstatedir}/%{var_base_dir}/%{name}
amanda.spec:273:%attr(02770,%amanda_user,%amanda_group) %dir /var/log/amanda
amanda.spec:274:%attr(02770,%amanda_user,%amanda_group) %dir /var/log/amanda/amandad
amavisd-new.spec:200:%dir %attr(770,amavis,clamupdate) %{_localstatedir}/run/clamd.amavisd
asylum.spec:110:%attr(2755,root,games) %{_bindir}/%{name}
...
How does it conflict with the sysusers stuff?
| @@ -1268,12 +1270,18 @@ rpmostree_compose_builtin_tree (int argc, | |||
| return FALSE; | |||
| gboolean changed; | |||
| if (!impl_install_tree (self, &changed, cancellable, error)) | |||
| return FALSE; | |||
| { | |||
| self->failed = TRUE; | |||
There was a problem hiding this comment.
Do we want this for rpmostree_compose_builtin_commit and rpmostree_compose_builtin_install too?
The problem is a system user with a differently named group; not a file owned by a different user and group. Which the current test conflates, but I think it would work to just drop the |
Ahh gotcha.
Yeah that's totally reasonable. I just didn't want to lose testing of user != group files. |
Sysusers doesn't really support this, and no RPMs do it.
cgwalters
force-pushed
the
sysusers-prep
branch
from
6a74406
to
ac81cb0
Compare
|
Updated this one. |
|
⚡ Test exempted: pull fully rebased and already tested. |
Will be used by the sysusers code to pass down a pipe that the child writes to. Closes: #1763 Approved by: jlebon
We should only be calling into GIO once we've set our preparatory environment variables etc. Not fixing anything today that I know of, just noticed while looking at the code. Closes: #1763 Approved by: jlebon
Sysusers doesn't really support this, and no RPMs do it. Closes: #1763 Approved by: jlebon
A few bits extracted from #1679